Traffic Light Series: Missing Data

Five times missing data changed the world

When you’re talking about using your data to run your organisation, the smallest amount of missing data can be a red light. Here’s five times that missing data changed the world as we know it.

1. A global pandemic is stopped in its tracks

When Hans Rosling arrived in Liberia during the start of the Ebola outbreak in 2014, the world was on the precipice of an epidemic. The disease was being treated aggressively in areas where it reared its head, but somehow it kept finding its way back into the population.

Rosling, one of the world’s most famous statisticians at the time, turned to the data. He saw the list of villages all reporting 0 cases. But Rosling was used to working with real-world data and knew something was wrong, because the list contained every village in the country, but he’d never seen all the villages in a country do anything at the same time. He realised that all those nice clean zeroes in the data weren’t villages reporting they were free of Ebola, they weren’t reporting at all.

He changed the zeros to correctly read as missing, and immediately medical scouts started visiting places that had been skipped. They found multiple unreported cases buried deep in remote parts of the country and quickly treated them, keeping the lid on a global disaster.

An aerial photo of the Gabba stadium in Brisbane

2. An Excel error breaks capitalism

In 2008 the world faced a financial crisis like never before. In the fallout of the crisis, politicians looked toward economists to pitch their solutions, and in 2010 two researchers provided them an answer that sounded right - austerity. Reinhart and Rogoff argued that high national debt was a trap that stopped countries growing, and that countries with no debt were more likely to survive economic woes.

Their research led to governments ruthlessly defunding anything that could be justified as a luxury - health, education, arts, infrastructure, and social services were all decimated. Every time protests got too loud the politicians would look sad and point to Reinhart and Rogoff’s paper, telling voters they were only doing what was economically rational, and thus best for all. Their hands were tied.

It was three years later that a US Student took a look at the original data, and noticed something. Thanks to the drag-and-drop nature of Excel, the researchers had made an error by only partially copying a formula across rows. Their data was missing for five countries.

When the formula was copied correctly, the analysis of the paper completely reversed - rather than cutting services, the fastest way out of trouble would have been increasing public funding across the board. The paper that had motivated most of the western world to cause generational pain to their citizens was retracted.

3. My graduation year is ‘Harvard’

Harvard has been through a bit of a rough patch with a couple of high-profile retractions, but none have been more ironic than the fraud conducted in the Psychology department’s Honesty research program.

The theory was compelling - the idea that tiny ‘nudges’ could massively alter how honest people were when nobody was watching. For example, one study found a massive difference in how honest people were on a form when they were required to sign it before giving the information, as opposed to after. Another study then extended the idea, showing that people felt more ‘dirty’ when asked to consider the ways in which they did not meet their obligations, compared to how they felt when thinking about their progress toward their goals.

It all unraveled in 2023, when a group of statisticians on the site ‘Data Colada’ published their own analysis of the honesty data. They realised that a number of participants had listed the start of their degree as ‘Harvard’ - the ineligible entry prevented reviewers from seeing that the true data (which should have been a year) was missing. Once they corrected this, they noticed that all the entries with missing years had results in favor of the researchers, and if these were removed the results flipped completely. They went through other studies published by the group and found the same thing - different types of ineligible data that should have been recorded as missing, always pushing the study in the researcher’s favour.

The cases around whether the issues constituted fraud have only recently been concluded, with Gino being reported as having left Harvard in June 2025.

4. Missing holes, and the planes that didn’t come back

A diagram of where returning planes in World War 2 had been hit.

As with many sciences, World War 2 saw a number of giant leaps in statistical analysis as everyone tried to gain every possible advantage. One of the most famous was the work of Abraham Wald in the US Airforce.

The air battles during the war were brutal, and each tour resulted in fewer planes returning. Those that did return were riddled with bullet holes. When they landed, analysts carefully recorded the position of each hole to try and improve the plane’s armor. However, their approach of covering the spots with the most holes was not improving their results at all.

At some point Wald realised that they were missing the critical data - where the holes were on the planes that didn’t return. With that understanding everything reversed. The planes that they had good data for were the ones that were hit in places that didn’t affect flight, so the armor needed to go everywhere there weren't holes. If you got hit anywhere they had a record for, your plane came back, but if you got hit anywhere else you didn’t. The change made a revolutionary difference in the protection of US soldiers, saving countless lives and millions of dollars.

5. Missing the giant hole in the Ozone layer

In the 1980s NASA was using satellites to monitor the state of the ozone layer as part of their regular scientific research. Satellites often record unusual data, and when you’re dealing with readings of things taken from space, you expect to see a wide range of numbers.

This led researchers to set some thresholds of when data would be considered ‘acceptable’. Specifically, they set a daily threshold based on their initial readings, and sent their sensors up to space. But British scientists on the ground were pointing their own sensors at the ozone layer and seeing the numbers plummeting, showing the hole appearing and growing. For five years the two groups bounced back and forth, with all researchers becoming increasingly convinced their instruments must be wrong. It was only once someone examined the data that had been auto-filtered from the NASA reports that they realised the horrifying truth - CFC gasses had dissolved the most critical part of the atmosphere which we rely on for survival.

In 1986, NASA confirmed their instruments were working fine, as were the British researchers. The world reacted accordingly. In one of the most inspiring movements of the modern era, the world stopped using CFC gasses in daily life, and the hole began to heal.

If all this makes you wonder whether missing data might be a problem for your organisation, you can visit our Laboratory for a free data health check, or set an intro call with the button below.

Book an intro call